摘要
提出了利用人工神经网络重构区域电离层临界频率分布的方案。对太阳活动高年和太阳活动低年白天的电离层临界频率分布进行了重构 ,讨论了站点分布对重构的影响 ,利用境外的历史资料对重构结果进行了验证。研究表明 ,拉萨站和重庆站对重构影响较大 ,兰州站和乌鲁木齐站对重构的结果影响较小。
A method for reconstruction f0F2 using artificial neural network (ANN) over limited region is described. Reconstructed contours were obtained both in high solar activity and low solar activity. At last the observed station's influence on the reconstruction was analyzed, and the reconstructed result was validated with the other overseas stations data. The result shows that Lhasa station and Chongqing station influence the reconstructed result more serious than Lanzhou station and Urumchi station.
出处
《电波科学学报》
EI
CSCD
2004年第6期742-747,共6页
Chinese Journal of Radio Science
基金
预先研究项目 (4130 70 310 2 )
关键词
人工神经网络
临界频率
重构
BP算法
Algorithms
Frequencies
Neural networks
Reconstruction (structural)
Signal processing
Solar radiation
